Sorry to hijack you: I have some questions regarding current weather models:
I am personally not interested in predicting the weather as end users expect it, rather I am interested in representative evolutions of wind patterns. I.e. specify some location (say somewhere in the North Sea, or perhaps on mainland Western Europe), and a date (say Nov 12) without specifying a year, and would like to have the wind patterns at different heights for that location say for half an hour. Basically running with different seeds, I want to have representative evolutions of the wind vector field (without specifying starting conditions, other than location and date, i.e. NO prior weather).
Are there any ML models capable of delivering realistic and representative wind gust models?
(The context is structural stability analysis of hypothetical megastructures)
I mean - you don't need any ML for that. Just go grab random samples from a ~30 day window centered on your day of interest over the region of interest from a reanalysis product like ERA5. If the duration of ERA5 isn't sufficient (e.g. you wouldn't expect on average to see events with a >100 year return period given the limited temporal extent of the dataset) then you could take one step further and pull from an equilibrium climate model simulation - some of these are published as part of the CMIP inter-comparison, or you could go to special-built ensembles like the CESM LENS [1]. You could also use a generative climate downscaling model like NVIDIA's Climate-in-a-bottle, but that's almost certainly overkill for your application.
take a look at all the roofs next winter, if its anything like the other side of the canal, you'll see that the average roof coverage is substantially less than 1/3.
nuclear energy still causes a lot of prompt heating
other forms of renewables could generate electricity while cooling the planet.
a super chimney (perhaps suspended with balloons) piercing the tropopause and carrying either air in open or closed loop fashion, or a "refrigerant" (not necessarily a harmful one, could just be moist air, or any other medium of thermal exchange, like a gravity assisted heat siphon) in a closed loop could generate power while cooling the planet, it would also be base load given the large temperature difference between surface level and tropopause (which persists day and night, summer and winter). Obviously this can also be used to desalinate sea water by freeze desalination.
as soon as such technology takes off and multiple blocs make use of such technology, they will probably even get into arguments about how long or what fraction of the time each nation state is allowed to generate power this way (arguing it was our Western excessive CO2 consumption to which we have to thank this excess heat availability, and India countering that we should take into account their proper share of excess CO2 due to the underground coal mines that have been burning uncontrollably for decades on end, etc...) to the point of nation states attacking each others superchimneys.
What I don't like when people bend over backwards to hype epigenetics:
1. the lack of historical context and opinion lock-in once expressed: before the human genome project many geneticists thought the number of "genes" (yes its an oversimplification) would be much much larger than what was eventually discovered to be the case. It was a shock to many biologists, that there were just about tens of thousands of genes in the genome, basically "a handful" in terms of control theory. The miracle of life started to look marginal and banal, just like many protested the earth not being the center of the solar system...
2. no highlighting the difference between single-cellular species and multi-cellular species: generational memory effects are obviously observable for single-cell organisms. For multi-cellular organisms, like humans, the concept of generation is ambiguous: is one talking about cellular generations (fertilized egg cell, dividing and the daughter cells dividing again, ...), or about organism-level generations (human parents of human children)? It becomes immediately apparent that the cell lineage from fertilized egg cell, to either sperm cell of the son or egg cells of the daughter, would involve lots of divisions, and the final sperm or egg cell would only have access to vague general variables: blood sugar levels, temperature, some hormonal levels, ...
3. Just like one can not truly study physics, without learning mathematics, and then formulating claims and observations mathematically, to truly study biology and the homeostasis it implements, you need systems biology: a mathematical description of biology. To understand multicellular organisms, one needs to understand the concept of cell types mathematically, and to summarize it in natural language cell types are the stable attractors forcing the cell contents to return to the closest state of the cell type. To make a rough analogy (cell/human, cell type/profession) then in contrary to humans, cell's don'
t have any memory which is independent of their cellular content. So apart from the content of the cell, a cell is amnesiac. Imagine humans without memory, but upon seeing the room in which they work, they can constantly re-understand what role they perform, and any deviation from what is ingrained in your local DNA copy of the genome is responded to. If you find yourself in a room with ovens and lots of dough, and some of the ovens have bread, but for some reason there's a cop's badge on the table, then you know you are a baker, and you throw out the cop's badge. If you are in a special car with red and blue lights, and you are behind the wheel, but there is for some reason an oven sitting on the passenger seat, then you are a cop, and you take the oven out of your police car.
Once you understand how cell types implement the memory necessary so that a neuron in your brain doesn't start to behave like a skin cell on your anus or vice versa, you understand the ridiculous proposition of epigenetics, a quest for the holy grail of all the missing information that the human genome project failed to find... Euhm well sure there is some modulation of transcription by histone modifications etc... but all of that can be modeled in the same language of reactions that is currently used to model Gene Regulatory Networks, with Gillespie simulators etc. Instead of wishing to vindicate ones old (and wrong) forgotten statements from before the human genome project due to psychological lock-in effects, it would be more productive to point out that in practice a lot of known useful data is ignored, so why not first make use of information we know exists before ingraining vague ideas about epigenetics in the next generation of students? stop ignoring the promotor regions and consensus sequences etc when sequencing genomes, there is a wealth of information to be had there, and personalized DNA-driven medicine will never take off until these are by default sequenced as well, as they directly relate to transcription rates! you know, good old classical gene regulatory network data.
1. I am confused, did copyright holders not amused by archive.today etc, intentionally serve CSAM material when they detected a visitor was in fact archive.today scraping one of their pages? It seems they are on the hook for more than just "inaccurate reporting of CSAM materials".
2. Is it a legally allowed tactic for copyright-luvva's to intentionally seek out CSAM content online, and then submit those URL's to sites like archive.today? Which entity is at greater legal peril, the one that aids the distribution of CSAM materials by intentionally having a site like archive.today archive CSAM content, or archive.today unintentionally being tricked into archiving CSAM content?
3. Everyone has traumas, of one kind of another. Each deals or tries to deal with them in their own way. Suppose a victim of crimes (still unpunished) finds or is informed of the presence of evidence online, and suppose this victim (regardless of how representative) finds the preservation of this evidence more important than the humiliation associated with it, how (in)just are laws that blanket suppress CSAM material? To give a more vigorous example: imagine you were raped by some no-yet-fallen UK nobility, and you are made aware of the presence of this evidence on some royal FTP server (or whatever), and you succeed in having archive.today "notarize" this evidence (independently from legal channels, since theres a suspiciously low amount of nobility being convicted, in contrast to your personal experience). These rules for supressing CSAM can be wielded as a sword precisely against those who fell prey to perpetrators...
It didn't sound like that at all, it just says DNA samples are collected and stored. The implication that such DNA in such samples can be sequenced after the fact is not novel at all, every time DNA is sequenced, it is first collected.
A collection of cars is also a collection of steering wheels, a collection of tires, a collection of seats, a collection of engines a collection of seats, ...
as blood contains white blood cells, and these cells tend to contain DNA, yes a collection of identified blood samples is also a collection of DNA (molecules).
A DNA collection doesn't need to have been sequenced to qualify as a DNA collection.
On the other hand blood samples degrade over time depending on how you store it. This makes DNA sequencing more difficult and/or impossible. Presumably the ROI (in a non-dystopian society) of storing those sample long-term doesn't make sense, especially if the primary usecase is screening for diseases (a random PDF from the Association of Public Health Laboratories says biomonitoring/biothreat samples are stored 1 year https://www.aphl.org/AboutAPHL/publications/Documents/ID_Spe... ).
So yes a collection of blood sample is technically also a collection of DNA sequences, but it has an expiry date (a short one compared to the lifespan of an individual!) contrary to a DNA sequence that's pure data.
This is wrong, the sample efficiency of DNA collection, amplification and sequencing technology has gone way up, to the extent that for example bubonic plague could be identified from DNA samples in the dental plaque of skulls which are multiple centuries old.
Also your statement directly conflicts with the purported confirmed utility of law enforcement getting warrants to use said samples.
You don't think that it's easier to identify whether something is bubonic plague than it would be to identify whether two DNA samples are related? Especially when one of them is aged such that it's not even a whole sequence anymore?
Yes, our ability to manipulate and read DNA has increased significantly in the past 40 years. But you can't create data from something that isn't there or has been corrupted beyond recovery.
And as far as the efficacy of police goes, I don't think that a warrant is sufficient to prove that there's confirmed utility in getting these samples.
I am personally not interested in predicting the weather as end users expect it, rather I am interested in representative evolutions of wind patterns. I.e. specify some location (say somewhere in the North Sea, or perhaps on mainland Western Europe), and a date (say Nov 12) without specifying a year, and would like to have the wind patterns at different heights for that location say for half an hour. Basically running with different seeds, I want to have representative evolutions of the wind vector field (without specifying starting conditions, other than location and date, i.e. NO prior weather).
Are there any ML models capable of delivering realistic and representative wind gust models?
(The context is structural stability analysis of hypothetical megastructures)
reply